Group Formation
Open in the Padhaiko app1 Group Formation
Project I (ENCT 413) is the first phase of the B.E. Computer Engineering major project in the seventh semester (2 credits). Unlike the sixth-semester Minor Project, which ends in one semester, Project I covers problem formulation, literature review, concept note, proposal defence, initial development, and a proof of concept, setting up the foundation for full implementation and final evaluation in Project II (eighth semester). Every group begins with formal group formation, supervisor assignment, and track selection.
Project I and Group Formation Rules
Under the official Institute of Engineering syllabus, students form project groups of at most three members working under an assigned department supervisor (plus an external prospective mentor when pursuing the internship-based track). Each group chooses one of three official tracks—research, product, or internship-based—and carries out an engineering problem either by extending its sixth-semester minor project or by proposing a new computer engineering topic.
1.1 Team Size, Complementary Roles, and Group Charter
Because a major project spans two semesters and involves design, implementation, testing, experimentation, and academic writing, a team of 2 to 3 members works best when skills complement rather than duplicate one another.
| Role Focus | Core Responsibilities | Typical Deliverables Owned |
|---|---|---|
| System / Backend & Data Lead | Architecture design, database/API or hardware integration, data pipeline setup | System diagrams, backend/firmware modules, dataset scripts |
| Core Algorithm / AI / Frontend Lead | Core model or algorithm implementation, UI/client interface, prototype logic | Trained models or core algorithms, UI screens, PoC demos |
| Integration, Testing & Documentation Lead | Test plans, benchmarking, version control hygiene, concept note, proposal & progress reports | Test logs, evaluation tables, Gantt tracking, formatted reports |
Before approaching a supervisor, a newly formed group should agree on four practical ground rules:
- Weekly sync schedule: Fix a consistent 30-minute internal meeting slot plus one weekly supervisor check-in.
- Version control and workspace: Set up a shared Git repository (with clear branch rules) and a shared drive for papers, datasets, and drafts from week 1.
- Workload parity: Track tasks on a shared board so coding, experiments, and documentation stay balanced.
- Continuity commitment: Confirm whether all members plan to continue the same project into Project II in the eighth semester.
1.2 Choosing a Project Track
Project I formally offers three execution tracks. Selecting the right track early determines what problem you frame, who advises you, and what proof of concept is expected by the end of semester seven.
| Aspect | Research Track | Product Track |
|---|---|---|
| Primary Goal | Investigate a technical hypothesis, propose or improve an algorithm/model, and benchmark results | Design and build a reliable, usable software, hardware, or hybrid system for target end users |
| Core Literature Work | Deep survey of peer-reviewed journals/conferences to locate a research gap | Survey of existing commercial/open-source products and user workflows to identify feature or usability gaps |
| Success Evidence in Project I | Baseline pipeline running on a dataset with preliminary quantitative metrics (e.g., accuracy, F1, latency) | Working architecture, core modules, database/UI prototype, or hardware testbed demonstrating end-to-end flow |
| Target End Product in Project II | Validated experiments, ablation studies, comparative evaluation, and potentially a conference/journal paper | Deployed or field-tested application/device with user evaluation, security checks, and full system documentation |
| Track | Supervision Structure | Best Suited For | Key Risk to Manage |
|---|---|---|---|
| Research Track | Department supervisor | Students interested in ML/AI, computer vision, NLP, security analysis, or algorithmic optimization | Dataset unavailability, vague hypothesis, or lack of reproducible baseline |
| Product Track | Department supervisor | Students aiming to solve a practical industry, campus, or community workflow problem with a deployable system | Over-ambitious feature list (scope creep) and weak engineering design |
| Internship-Based Track | Department supervisor + Prospective industry mentor | Students placed at a software/engineering organization working on a well-defined engineering module | NDA/confidentiality limits on code or data; mismatch between company tasks and Institute of Engineering academic milestones |
1.3 Supervisor Assignment and Theme Alignment
The department assigns a supervisor to guide topic refinement, monitor milestones, and review deliverables before submission. Groups may also extend their sixth-semester Minor Project or pick a fresh topic from core computer engineering domains such as:
- Artificial intelligence and machine learning (AI/ML)
- Image processing and computer vision
- Computer networks and cybersecurity
- Natural language processing (NLP) and speech/audio systems
- Embedded systems, robotics, and Internet of Things (IoT)
- Big data technologies, cloud systems, and web/mobile applications
Forming a balanced group and selecting a track for a Minor Project extension
Three seventh-semester students built a basic desktop attendance prototype using Haar cascades during their Minor Project. They want to continue together for Project I. Outline how they should structure their group, pick an appropriate track, and define their role split before meeting their assigned supervisor.
Audit team skills and past gaps: Student A is comfortable with Python and PyTorch, Student B has backend/REST API and database experience, and Student C excels at frontend UI and system testing. In their Minor Project, Haar cascades failed under poor lighting and masked faces, and attendance logs were stored only in local CSV files.
Select the track and justify upgrading the theme:
They choose the Product Track (with an embedded ML component): building a classroom-ready, edge-to-cloud attendance system using modern face embeddings, anti-spoofing liveness checks, role-based web dashboards, and real classroom lighting trials.
Draft a clear role matrix and operating charter:
- Student A: Face embedding pipeline, anti-spoofing model evaluation, and edge inference optimization.
- Student B: Backend API, PostgreSQL schema, synchronization between edge camera client and server, and security.
- Student C: Admin/teacher web dashboard, automated test suite, lighting benchmark logs, and documentation lead.
All three agree to meet Tuesdays/Fridays internally and maintain one shared Git repository.
The group forms a 3-member team on the Product Track, extends its Minor Project by upgrading detection, adding liveness checks, backend persistence, and real classroom validation, and assigns distinct technical ownership plus shared Git/meeting discipline before the first supervisor meeting.
Preparing an Internship-Based Track Group Charter
Two students secure a seventh-semester internship at a local fintech company working on transaction anomaly monitoring, while a third classmate is not interning there. Outline how the group should handle group composition, track selection, and mentor coordination.
Check group composition and track consistency: Under the internship-based track, every group member's contribution must be evaluable by the department and aligned with the project deliverables. Either the two students form a 2-member internship group (since maximum is 3, pairs of 2 are valid), or the third member joins only if the company formally includes all three on the same project with a shared scope and mentor sign-off.
Separate proprietary assets from academic deliverables:
The two students confirm with the fintech mentor that they can use anonymized or synthetic transaction logs for department demos, concept notes, and progress reports while developing the core rule-and-ML anomaly engine during their internship.
Align dual supervision: The group records both the assigned department supervisor (responsible for Institute of Engineering academic format, week-2 concept note, week-3 proposal, and semester-end evaluation) and the prospective industry mentor (responsible for domain guidance, technical validation, and mentor approval form).
The two students form a 2-member group on the Internship-Based Track (keeping the non-interning classmate free to join another group), secure written mentor agreement on anonymized demo data, and align both the department supervisor and industry mentor on Institute of Engineering's milestone schedule.
Viva-Style Questions: Group Formation
What is the maximum allowed group size in Project I (ENCT 413), and what happens if a group has four students who worked together in Minor Project?
Official size limit: According to the ENCT 413 syllabus, students must be organized into groups of at most three members under the supervision of an assigned supervisor.
Handling a four-member Minor Project team: A four-member team cannot continue together as a single four-member Major Project group. The students must split (for example, into two pairs of two, or one group of three and one member joining another team) and obtain departmental approval if one sub-team extends the Minor Project theme.
The maximum group size is three members. A four-member Minor Project group must split into compliant groups of up to three members before registering for Project I.
Differentiate between the Research Track, Product Track, and Internship-Based Track in Project I. How does supervision differ across them?
Purpose of each track:
- Research Track: Focuses on formulating a research question/hypothesis, designing experiments, improving or comparing methods, and analyzing quantitative results.
- Product Track: Focuses on engineering a usable, end-to-end software, hardware, or hybrid system addressing a real user or industry problem.
- Internship-Based Track: Executes an engineering or R&D problem inside an external organization as part of an internship.
Supervision structure: Research and Product tracks are guided by an assigned department supervisor. The Internship-Based track requires both the department supervisor and an external prospective mentor from the host organization, who must approve the topic and monitor technical progress.
The Research Track produces validated experimental findings, the Product Track produces a working user-focused system, and the Internship-Based Track executes a supervised industry problem. Research and Product tracks work under a department supervisor, whereas the Internship-Based track adds a prospective industry mentor.
Under what conditions is a group permitted to extend its sixth-semester Minor Project into Project I? What would cause an evaluation panel to reject a Minor Project extension?
Permitted extension conditions: The syllabus allows extending the Minor Project when the team introduces substantial new engineering or research depth—such as new algorithms, real-world dataset collection, scalable architecture, hardware integration, or rigorous performance evaluation across two semesters.
Reasons for rejection: A panel rejects an extension if the proposal merely repeats Minor Project objectives, adds trivial cosmetic changes (like UI color tweaks), or lacks enough technical complexity to justify a 2-credit Major Project Phase I and subsequent Project II.
Extending a Minor Project is permitted when the group adds substantial new research or engineering scope (real datasets, improved algorithms, full-system integration, or rigorous evaluation). Cosmetic upgrades or repeating Minor Project deliverables lead to rejection.
During a viva, an examiner asks a group member about the backend database schema, and the student replies, "I only worked on the frontend UI." Why is this a problem, and how should a group prevent it from Week 1?
Why it hurts evaluation: Project I evaluates each student's engineering understanding of the complete system, not just isolated task completion. A response showing total ignorance of another module suggests poor collaboration, weak system integration, and uneven engineering rigor.
Preventive group practices: Define clear module leads for accountability, but hold weekly internal walkthroughs where each member explains their code/design to the others, review pull requests across teammates, and co-author the architecture section of the proposal.
Specializing in one module does not excuse ignorance of the rest of the system, because examiners assess every member on the overall architecture and integration. Groups prevent this through weekly cross-member walkthroughs, shared code reviews, and collaborative report writing.
A group on the Internship-Based Track cannot share its host company's production source code or customer database with the university. How should the group plan its Project I deliverables so it satisfies both the company and the department?
Agree on boundaries before proposal submission: Discuss confidentiality and intellectual property with the prospective industry mentor and department supervisor before week 2.
Use anonymized or synthetic substitutes for academic review: Design the system architecture diagram at an abstract/module level without leaking trade secrets, use anonymized or synthetic sample datasets for proof-of-concept benchmarks, and demonstrate core algorithmic or architectural contributions rather than proprietary credentials or client records.
The group should secure early agreement from both the industry mentor and department supervisor to present abstract architectural designs and evaluate its proof of concept on anonymized or synthetic datasets, keeping proprietary data confidential while still proving technical competence.